• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

LDO or discrete switching reg's in series?

Status
Not open for further replies.

causalitist

New Member
is there a way i can put say 3 Linear voltage regulators in series to step down about 3 times the Vin rating of one?
can i do this with a little to-220 switching regulator like
LM2575HV ? (really cool product line i didnt know about!)

say i have 110vdc . and want 15vdc.. could i put 3 40v LDO in series? (that each give Vo =5v or Vo=15v)

i know there are LDO's that can take 125vin .. but im curious if i could series 3 rated at lower volts.. wouldnt do it because of heat, but ive been wondering..

and then also, same 110v to 15v : can i put 2 or 3 LM2575HV switching reg in series?

i need like 150ma... but this is more about if i can do it, as opposed to me asking you to give me a schematic solving all my lifes problems hahaa



crap you dont have to read:
(im looking for a 1st stage in stepping down my 110vdc power rail volts in my 50amp 110v IGBT pwm motor drive, for a bicycle. 110v is lithium polymers. after i can get the volts to around 16-20v, then i will clean/regulate/isolate... and "drive the crap outa that igbt!! wooo!" ... sorry.. hahaa)(yes, ive considered a good isolated dc-dc for the whole sha-bang, but trying to save cost and see if i can get it lower cheaper and use a 15v to +15v, -10v dc-dc)

:confused:d:mad:≈ζ
yes... u heard me! (integral of confusion) d(anger) kinda equals dampening factor
 

Sceadwian

Banned
I'm not so sure three switching regulators in series would get along so well, they're going to be out of phase and who knows what kind of interactions would go on. You could try it but you're out the cost of the chips if it doesn't work. Have you thought about using a very simple oscillator to feed a small step down transformer (it'd be small because you could make the frequency high) then re-rectify it into the cheaper SMPS? That's how CFL bulbs do it to keep component size/cost down, although their transformers are for current limiting not voltage stepdown, same theory though.
 
Last edited:

ronsimpson

Well-Known Member
Most Helpful Member
Why "clean/regulate/isolate"? Drive the motor with the output of the PWM.
What do you have? 110 volt battery? Why so high voltage?
What do you need? 15V? +/-15? +/-10V? ?????
What motor?
 

causalitist

New Member
Why "clean/regulate/isolate"? Drive the motor with the output of the PWM.
What do you have? 110 volt battery? Why so high voltage?
What do you need? 15V? +/-15? +/-10V? ?????
What motor?
the motor is PMDC 130vdc, "3.5hp" McMillan electric co. model s3480b3292 25amp .. other than that i measured 3 Ohms, so like 444Amps stall.. i measured 2mH inductance (-+20%) .. but that doesnt matter at all and is not what i asked.

why such high voltage??? so it pulls more amps.. why else.
and for 10,000watts of power, 148v looses 4449*R power loss..
but 36v has 160,000*R lost! thats why.

ronsimpson - im not cleaning the pwm! im cleaning the battery voltage that has 10kw of dv/dt so i can power chips to MAKE pwm. .. to drive the motor.

the negative voltage is for the igbt gate.. i suppose you are not familiar with igbt's. the square wave to the igbt gate is +15v, -10v, for proper turn off.
i dont need that anymore.. i bought a 12amp gate driver that has its own dc-dc with +15v/-12v output and optoisolation..
 

causalitist

New Member
anyway. I DONT HAVE a problem that needs solving. i know how to do what i need to do. i want to know if i can do something JUST TO KNOW.

I want to know if i can put multiple LINEAR VOLTAGE REGULATORS in series to allow for a higher Vin. does anyone know the answer to this exact question?
 

ronsimpson

Well-Known Member
Most Helpful Member
If I understand you right; you are talking about the power to run the PWM chip and not the 50 amps to run the motor. Usually the PWM, running from a high voltage source, lives from a boot-strap supply. See application note for UC3842. (or any power line PWM) Use a 100k resistor from +110 volts to Vcc of PWM. This is not much current so the PWM can not run but it will start! When the PWM starts it supplies power to keep it's Vcc up.

When a battery is involved linear regulators are not used.
 

Sceadwian

Banned
Mr RB I'm not sure why you said no, you can definitely put linear regulators in series to increase the VIN, however I'm sure there may be problems that come up doing it, like oscillations for one, and generally.
 
Last edited:

Oblong

New Member
There is nothing wrong with putting regulators in series. As long as you don’t exceed any of the manufacturers specs (Vin max and( Vin max – Vout ) and power dissipation).

You likely do it all the time when you power your circuits off your bench top PSU.

SMPS’s are cascaded if the duty cycle gets to low as well.

But in your case the best choice would be to use a topology off the bat that would meet all your stated requirements such as a flyback or forward converter. The winding ratio could be used to drop most of the overhead voltage and it would also give you the required isolation. For 150mA at 15V ( 2.25W) About 3VA You may even be able to use an off the shelf pulse transformer. I’ve used them for small isolated power supplies.

There are plenty of HV controllers for NON-isolated bucks as well On semi has some in there NCP series as well as ST Viper series and others .

For your linear regulators you will have to dissipate over 14W!
Where as an SMPS see pic below Viper design software.

Input 80 -to 140VDC output 15Vdc @150mA. 82% conversion efficiency.

 
Last edited:

crutschow

Well-Known Member
Most Helpful Member
Mr RB I'm not sure why you said no, you can definitely put linear regulators in series to increase the VIN, however I'm sure there may be problems that come up doing it, like oscillations for one, and generally.
A problem would be on startup, when the total input voltage would instantly appear across the first regulator before its output capacitor was charged (since most linear regulators need an output and input capacitor for stability). That would blow the first regulator. The rest would then go like dominoes.
 

Sceadwian

Banned
Use a bigger input capacitor than output. Electrolytics have a nice amount of ESR built into them which will cause a 'soft on' state.
 
Last edited:

crutschow

Well-Known Member
Most Helpful Member
Use a bigger input capacitor than output. Electrolytics have a nice amount of ESR built into them which will cause a 'soft on' state.
You would need a power resistor in series with the first stage input capacitor, which would be okay since you need to drop a lot of voltage anyway.

But the capacitor ESR won't help that since it's internally in series with the capacitance.
 

Mr RB

Well-Known Member
Replies :eek:

He asked if you could string a lot of 3pin regulators to allow you to exceed their INPUT voltage limits.

So if you string three 7805 regulators, in series, on a 120v dc supply... Just how do you parallel their outputs? :eek:

Cascading 3 pin regulators is ok if the total Vin < 40v... But putting them in series???
 

MrAl

Well-Known Member
Most Helpful Member
Good point. Just because the 'input' can take up to 40v that doesnt mean
that the regulator will 'drop' 40v when it is operating.
I would think that it would be VERY hard to predict the impedance of any of the
regulators too, which would affect how much voltage each one receives.

For example, with a 120vdc line going to the first regulator input, where does
the ground line for that first regulator go? Wherever it goes it has to be
80vdc or more ALL THE TIME or the first regulator goes pop, only it makes
a lot more noise than that :)
 

crutschow

Well-Known Member
Most Helpful Member
An LM317 regulator, for example is not directly connected to ground. It, in effect, floats on the resistor network used to generate the desired output voltage. Thus, if you can soft start them so that none will see more that 40V, you can connect them in series (each one's output feeding the next's input) with each one set to output say 35V below it's input.

Granted that may not be the best way to drop a high voltage to a low voltage, but it is feasible. And if you short the output, it may cause them all to pop since they will not all current limit at the same point.
 

MrAl

Well-Known Member
Most Helpful Member
Hi again,


Yes, and other problems which are harder to foresee may 'pop' up
(little pun intended there) :)

This would not be something i would build. I would go with a more
standard approach that will surely work. This way i wont waste time
or money on pooped out parts lost during the prototype period.
 

causalitist

New Member
wow. thanks guys. first of all you should know i am only in circuits 1, none of my classmates even know what a diode is.. so while ive learned a massive amount on my own, i'm missing certain key concepts.

there are like 6 posts with very relevant information. too many to reply to directly. many of the fundamental issues like instantaneous voltage ect i was anticipating.. but i obviously hadnt thought the actual hookup through.. they cant have series inputs and || outputs hahah took me 5 seconds after reading that to feel like an idiot!

i believe you all have assumed the application correctly.. im lowering my main (actually 148vdc lith-poly) battery voltage ("tapping it") to power everything beside the motor.. and with such high dv/dt the main 148v battery voltage is naaaasty.

i looked for days seemingly EVERYWHERE for dc-dc or linear solutions... seems like a premade(i really dont want to make my own, .. one learning experience at a time) dc-dc only come with max Vin of 75v...

i ended up checking out powerex, who actually makes the 1200v, 600amp igbt im using, and they sell exactly what i need: M57182N-315
(http://www.electro-tech-online.com/custompdfs/2009/05/m57182n_315.pdf)

and their gate driver seems to be the most rugged one stop shopping solution i have ever seen.

i will use the above HV to 15Vo DC-DC .... into one of their 12A, optically isolated/ dc-dc isolated/ +15v,-10v gate drivers. perfect.

i had searched digi, mouser, allied, farnel,newark .. not one had what i needed.

so i just wanted to see, out of curiosity, if there even was a inefficient/simpler way of dropping 133v while retaining 100-200ma. still seems like its not even possible besides an old fashion voltage divider with resistors the size of new Mexico.

i was curious because this huge motor is on a mountain bike, and my 148v, 4amp lithium polymers will only last around 3.5 minutes at 10KW.. so a handful of watts lost at a fixed amount per unit time is no big deal .. so i wanted to see if there was a simpler way besides dc-dc.. mostly for future knowledge since i found this powerex combo.

so i gather hooking a bunch of linear reg in CASCADE(sorry, we just started 2 port networks in school) IS doable, if input caps are huge(makes sense)... .. for each?
 
Last edited:
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top